Table 1 The values of super-parameter tuning
From: DeepUbi: a deep learning framework for prediction of ubiquitination sites in proteins
Super-Parameter | Preferred Setting |
|---|---|
Embedding length | 21 |
Batch size | 65 |
Maximum epoch | 30 |
Convolution blocks | ([2–6], 64, ReLU) |
Fully connected layer units | 128 |
Cutoff | 0.5 |
Dropout | 0.7 |
Learning rate | 0.01 with decay rate 0.95 |
Regularization | L2 |